@inproceedings{3bd0fe00fe22423cb2a2806e2574818d,
title = "An Embedding-Aware LSTM-Refined Feature Attention Network for Structured Cyber Attack Detection",
abstract = "An embedding-aware LSTM-refined feature attention network for structured cyberattack detection is presented in this work. The suggested model converts the input traffic into token-like representations by integrating continuous and categorical features using scaling and embedding layers, respectively. While values are refined using an LSTM layer to incorporate sequential context, queries and keys are generated using linear projections. These representations are aligned by an attention mechanism to highlight the most instructive aspects. The model's robustness is demonstrated by evaluation on the CIC-DDoS2019 dataset, which yielded a 93\% accuracy rate.",
keywords = "Cyber attack detection, Intrusion Detection System, LSTM Attention, feature embeddings",
author = "Akshat Gaurav and Gupta, \{Brij B.\} and Nadia Nedjah and Hsu, \{Ching Hsien\} and Chui, \{Kwok Tai\}",
note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 14th IEEE Global Conference on Consumer Electronics, GCCE 2025 ; Conference date: 23-09-2025 Through 26-09-2025",
year = "2025",
doi = "10.1109/GCCE65946.2025.11275464",
language = "English",
series = "GCCE 2025 - 2025 IEEE 14th Global Conference on Consumer Electronics",
pages = "589--590",
booktitle = "GCCE 2025 - 2025 IEEE 14th Global Conference on Consumer Electronics",
}